The Stereoisomerism of Complex Inorganic Compounds. XIV. Studies

The Stereoisomerism of Complex Inorganic Compounds. XIV. Studies upon the. Stereochemistry of Saturated Tervalent Nitrogen Compounds1. By John R...
0 downloads 0 Views 563KB Size
July 20, 1952

STEREOCHEMISTRY OF SATURATED TERVALENT NITROGENCOMPOUNDS

3535

postulated that these different straight lines were obtained because of the differences in double bonded character of the chelating groups. Such a difference in intercepts is not found for the corresponding phenanthroline-bipyridine complexes. The difference in double bonded character does not appear to be a factor in this case. Oxidation-Reduction Potentials.-The linear dependence of log kdiss upon log K , suggests that other relationships between various properties of the series of compounds studied might add to the understanding of related systems. Since the series studied included only the 5-substituted phenanthrolines, the question naturally arises as to the possibility of extending the observed generalizations to other related compounds. The extent to which this type of approach can be extended will ultimately determine their usefulness. Data for the oxidation-reduction potentials are given in Table I.’ The logarithms of the oxidation-reduction potentials may be plotted against log kdiss for the complexes and a straight line results. It has been observed that in many of the iron(I1)phenanthroline complexes a change in hydrogen ion concentration from 0.1 F to 1.0 F causes approximately a 0.034.04 volt change in formal potential. If one substitutes the expressions for the appropriate dissociation constants into the Nernst

equation, i t is possible to calculate the stabilities of the ferric complexes. The calculated values for 1,lo-phenanthroline agree approximately with those obtained by Lee, Kolthoff and Leussinglbfrom potentiometric measurements only if a l : l combination of Fe(Phen)8+f+ and H + is assumed. However, the assumption of a protonated ferric complex does not appear to be logical when this is not postulated for the ferrous complex. It does not seem reasonable, either, that the ferric ion with its higher charge would need an extra proton to ensure stability while the ferrous complex is apparently unaffected by hydrogen ion. I t is noted that if the data for the bipyridine complex are included on the graph of the oxidationreduction potential against log kdiss the resulting point does not fall on the line obtained for the 5substituted phenanthrolines. Work by Mellon and Hale on the 3-carbethoxy4-hydroxyphenanthroline complex with iron( 11) points to another apparent exception.* Although the oxidation-reduction potential would indicate that this latter complex would be more stable than any of the complexes studied here, i t was reported to be relatively unstable. I n these cases other factors which were unimportant in the 5-substituted compounds, may influence the stability of the complexes.

(7) G. P. Smith and P.P. Richter, “Phenanthroline and Substituted Phenanthroline Indicators,’’ The G. Frederick Smith Chemical Company, Columbus, Ohio, 1944.

WEST LAFAYETTE. INDIANA

(8) M. N. Hale and M. G. Mellon, THISJOURNAL, T l , 3217 (1950).

[CONTRIBUTION FROM THE

NOYES LABORATORY OF CHEMISTRY, UNIVERSITY

OF ILLINOIS]

The Stereoisomerism of Complex Inorganic Compounds. XIV. Studies upon the Stereochemistry of Saturated Tervalent Nitrogen Compounds’ BY

JOHN

R. KUEBLER, JR.,

AND JOHN

c. BAILAR,JR.

RECEIVED APRIL9, 1952 The reactions of N,N-diethylglycine and N-methyl-N-ethylglycine with cobalt(111) and platinum( 11) have been studied. A satisfactory synthesis for potassium dinitro-(N-methyl-N-ethylg1cyine)-platinate( 11) has been developed, That an asymmetric nitrogen atom exists in this compound was demonstrated through its resolution by fractionation with l-quinine and treatment with optically active quartz powder.

The problem of resolving into optically active antipodes compounds containing properly substituted nitrogen atoms has attracted interest for many years. Early investigators worked chiefly with organic compounds, and many resolutions of quaternary ammonium salts and tertiary amine oxides have been reported.2 I n 1924, Meisenheimer and his co-workers8 claimed to have shown the existence of the required four optical isomers of sarcosine-bis-ethylenediaminecobalt(II1) chloride, [Co enzsarc]Clz,4by frac( 1 ) The work reported in this article was taken from the doctorate thesis of J. R. Kuebler, Jr., 1951. (2) R. L. Shriner, Roger A d a m and C. S. Marvel, “Organic Chemistry;’ Ed. 2, Vol. 1. H. Gilman, ed., John Wiley and Sons, Inc., New York, N. Y., pp. 413-418. (3) J. Meisenheimer, L. Angermann, H. Holsten and E. Kiderlen, Ann., 488, 269 (1924). (4) Abbreviations used herein include dien dietbylenetrinmine, en ethylenediamine, amac N,N-diethylglycinnte ion, AMAC N-methyl-N-ethylglycinateion, glyc * glycinate ion, quin quinine ion, sarc sarcosinate ion, stry strychnine ion, d dextro-

-

-

-

-

=.

-

tional crystallization of the bromocamphorsulfonate salt. The interpretation of this work is complicated by the fact that the activity due to the cobalt ion is very large relative to that of the nitrogen atom and by the fact that when the resolving agent was removed, no nitrogen activity was detectable, but only that due to the cobalt ion; both of these factors cast doubt on the validity of the work. In addition, Basolo5 was unsuccessful in an attempt to repeat Meisenheimer’s procedure. Later, Manne prepared tetrachloro-(diethylenetriamine monohydrochloride)-platinum(1V) monohydrate [Pt(dien.HC1)C1~]~H~O1 but was not able to resolve it. In these previous attempts to resolve inorganic rotatory, and I = levorotatory. D- and L- represent the observed rotatory behavior of the complex ions. All rotations were measured at the D line of sodium. (5) F. Basolo, Thesis, Doctor of Philosophy, University of Illinois, 1943. (6) F. G. Mann, J . Chcm. Soc., 466 (1934).

complex coiiipounds coritainirig an asynimetric nitrogen atom, one of the four groups attached to the nitrogen was a hydrogen atom. I t is iiow bclieved that in such cases the hydrogen atotii is labile enough to ionize to a slight extent, thus causing continuous racemization of the optical forins. I t seemed reasonable that this process occurred in the above complexes, especially since Block and Bailar? were able to treat the [ h e n z ]4 ion with one equwalent of base to forni the [.Au(en)(en-Hj]+2 ton. (“en-H” represents a molecule of ethylenediamine from which one hydrogen atom has been completely lost.) Consequently, S-methyl-N-ethylglyei~ewas selected as a coordinating agent. i t was realized that, although such complexes iiiight be optically stable, the tertiary nitrogen atom would have a decreased tendency to form a coordiiiate botid with a iiietal ioii. However, the possibility of chclatioti through the carboxyl group should largely overcome this disadvantage. In addition, i t seeiiied desirable to synthesize a conipound in which the type and configuration of the other ligands offered no possibility of optical isoiiierism, so that the activity, if found, would necessarily be due to the nitrogen atom alone. % W e studying the synthesis of complex compounds, S ,N-diethylglycine was used as a “pracLice” coordinating agent since starting materials for i t 5 preparation were inore readily available than t h e for the methyl-ethyl compound. Clearly, thii: coinpound differs From ,“;-n~ethyl-li-ethylglycine only in its inability to form an asynimetric nitrogen atom upoii coordination to a metal ion, and subsequent results ha\-e shown that the two amino ‘tctds react otherwise in the same manner. First, the reaction of N,K-diethylglycine with several cobalt coiiipounds was investigated. Attcniptcd replacement of the chloride, ammonia, or sulfite, respecti\. -, in cis- [Co(NH3)dCl2]I, [Co(SOz)i(KHn,jLl cl [CO(?;H~)~(SO&] - by the ~tiiiino acid was unsuccessful. Furthermore, the dtiiirio acids, glycine, sarcosine and X,S-diethylglycine, were made to react with cobaltic hydroxide by tlie method of 1 , q and Winkler,6 to determirie the cffrct of substitiltioii of the nitrogen .Ltotii on ~0111Illex stability. (;lycinc foriiir the very stable tvis coiiiplex, [Co(glycJ $1 91120. Evidence of an analogous, but less stable, sarcosine complex was obtained, although the compound was never isolated 111 pure form. S ~S-I)iethylglycinc, however, showed no reaction. i t was concluded that the tertiary nitrogen, eve11 though aided by tlie possibility of chelation, was not capable of forming a stable cobalt -nitrogen bond. Xext, the use of platinum(I1j as a central ion was iiivcstigated. Even though the tendency for platiiiutIi-oxpgeri coordination is not very pronounced, d nuinber of compounds are known which contain >table platinun-oxygen bonds,9 and i t appeared probable t h d t ?j rrieth~l-S-ethplglyciticwould eoor d i nate. N,N-Diethylglycine allowed to react with potas 7 , €3 1’ RIock atid John C Baildr Tr

, 1351

.r

H Ley a n d l i Winkler Be1 iI

x

I H I S J O U R I . ~ L , 73, 4722

42, 3804 (19OY K i n g J Chcrit So< 1.338 (19181 K C I f r i i / i c s ihid II r, I K F ~ , L WR ~ ,46 ~i 1112)

siuni tetrachloroplatiiiate(I1) to form bis-(N, N-diethylg1ycine)-platinum (11), [P t ( a m a ~ ) ~ ]No . attempt was inade to prepare the analogous potentially active coinpound, [Pt(AMAC)2], since it would be a non-electrolyte and not amenable to the conventional method of resolution by salt forniation with an optically active compound. In addition, it would contain two identical asymmetric nitrogen atoms, and thus could conceivably occur in an iiiactive, meso form. From the above results, i t was concluded that a resolvable platinuni(I1) compound could be prepared from a platinum (I1j complex containing two halogen atoms and two other negative or neutral groups that are so strongly coordinated to the platinum(I1j ion that there is no possibility of their being displaced by the amino acid. Potassiuni diiodo-dinitro-platitiate(L1) dihydrate, K2[Pt Jh’02)212].2€120,was prepared by a modificatioti of the method of Nilsctnio and its cz\ configuration was denioristrated by its rcady conversion to [ Pt (X 02’)2en1. A solution of li2[Pt(NO2)2I2].BH20 was shake!: with excess silver oxide, the silver oxide and silver iodide were filtered off and the amino acid complex was obtained by treating the strongly basic filtrate with the amino acid nitrate. Both K[Pt(NO&amac] and K[Pt(N02)2AMAC]were readily prepared in this manner; however, they were so hygroscopic that it was necessary to convert them to the strychnine salts in order to obtain reliable elemental analyses. The nitrate of the amino acid was used rather than the hydrochloride, because the nitrate ion, being a poor coordinating agent, could riot compete with the tertiary nitrogen atom for a place in the coordination sphere of platinum(I1). The resolution of potassium dinitro- (N-methylS-ethylglycinej-platinate(I1jwas achieved in two ways, by fractionation of an optically active alkaloid salt, and by adsorption on optically active quartz powder. Cinchonine, brucine and strychnine salts were found to be too insoluble to be of IISC. The I-quinine salt wa5 too insoluble to give an observable rotation, but it was found possible to convert it to the potassiuni salt, KIPt(XOzjAMAC], several fractions of which exhibited small rotations which fell to zero in the course of five to six hours. The fact that all the fractions examined showed only levo rotation does not appear implausible, due to the niatiner in which they were obtained. ’The quinine salt of the levo isomer was apparently the less soluble of the two and concentrated in the crystals as the solution was evaporated. The more soluble dextro isomer concentrated in the solution, but as the evaporation was slow and the fractions were reinoved only every 12 hours, the dextro isonier had ample time to racemize. Early investigatorsi1 observed that when solutions of racemic mixtures of certain complexes were stirred or shaken with d- or I-quartz, the solutions showed optical activity. This was attributed to a preferential adsorption of one isomer by the optically active quartz. However, the mechanism inIO) I F Nilson, J pvakl CAcm , [ 1 1 ] 21, 172 (1880) ( 1 1) I< I snchidd, hl Kobaydshi and .4 Nakamurd J CiieiH S o

July 20, 195%

STEREOCHEMISTRY OF SATURATED 'TERVALENT NITROGENCOMPOUNDS

3537

glass, washed with alcohol and ether, and recrystallized from water. Anal. Calcd. for [Pt(amac)~]: C, 31.65; H , 5.32; S , 6.16. Found: C,31.40; H, 5.28; K , 6.17. Preparation of Potassium Tetranitroplatinate(11) Dihydrate, KL[Pt(NO&] ~2Hz0.-Potassium tetrachloroplatinate(II)'3 was converted t o the tetranitro compound by the method of Vezez." Two and two-tenths grams (0.005 mole) of potassium tetrachloroplatinate(11) was dissolved in 10 ml. of water and heated on a steam-cone. To this was added a solution of 3.2 g. (0.034 mole) of potassium nitrite in 5 ml. of hot water. The deep red color of potassium tetrachloroplatinate(I1) immediately changed to pale yellow, and upon cooling the solution in an ice-bath, white needles of K2[Pt(NO& .2Hz0 separated. A second crop was obtained by evaporation of the filtrate on a steam cone to one-half its volume, then cooling again. The crystals were washed with a small portion of ice water, then alcohol and ether and dried in the air. The total yield was 2.0 g., Experimental or 76%. Preparation of Potassium Diiododinitroplatinate( 11) DiPreparation of N-Methyl-N-ethylglycine Hydrochloride..2Hz0.-Six and five-tenths grams A modification of the method of Jones and Major'l was used. hydrate, K2[Pt(NOz)z12] was dissolved in 110 ml. The methylethylamine hydrochloride (Eastrnan Kodak (0.014 mole) of K2[Pt(NOz),]~2H20 Company) was purified by recrystallization from absolute of water and 110 ml. of alcohol was added. To this was alcohol and absolute ether. To 10.9 g. (0.114 mole) of added a solution of 3.65 g. (0.014 mole) of iodine in 230 ml. tnethylethylamine hydrochloride 25 ml. of 50% potassiutn of 50% alcohol. When the solution was heated 011 a steamhydroxide solution was added slowly from a dropping funnel. cone for one-half hour, acetaldehyde was evolved and the initial deep red color changed t o a yellow-brown. The The flask was heated on a steam-cone and the methylethylamine distilled into a flask packed in ice. The product was solution was concentrated t o a volume of about 80 ml. by heating on a steam-cone while air was passed over the surdried for 12 hours over potassium hydroxide pellets. To a solution of the methylethylamine in 50 ml. of abso- face of the solution. The small amount of black residue lute ether was added, dropwise, a solution of 12.6 ml. (0.114 which formed was removed, and the bright yellow solution mole) of ethyl bromoacetate in 50 ml. of absolute ether. A was evaporated to dryness in a vacuum desiccator over white precipitate of methylethylamine hydrobromide started phosphorus pentoxide. The yellow residue was recrystalto form at once and was filtered after one-half hour. An- lized from boiling, absolute alcohol. Yellow leaflets sepahydrous hydrogen chloride gas was bubbled into the filtrate rated from the solution upon cooling in an ice-bath, and were until precipitation of the amino acid ester hydrochloride was filtered and washed with cold alcohol. A second crop was complete. The precipitate was filtered, washed with ab- recovered from the filtrate by adding twice its volume of solute ether, and dissolved in water. The aqueous solution absolute ether. The over-all yield was 7.6 g., or 88%. I, 38.65. Found: was extracted three times with ether t o remove the last Anal. Calcd. for K2[Pt(NOz)~I~].2Hz0: traces of unreacted ethyl bromoacetate, and the extracts I , 38.85. Preparation of Dinitroethylenediaminepletinum(I1) Didiscarded. The solution was made alkaline with 5% soand Dinitroethylenediaminedium hydroxide solution and extracted eight to twelve hydrate, [ Pt( NOz)zen] .2H20, times with 1 5 4 . portions of ether. The ether solution was platinum(II), [Pt(NOl)nen] .-One and three-tenths grams was shaken one-half dried over anhydrous calcium sulfate. After distillation of (0.002 mole) of K~[Pt(N02)21z].2H20 the ether and unreacted methylethylamine, the oily residue hour with 100% excess silver oxide and 5 ml. of water. of amino acid ester was hydrolyzed with 50 ml. of 1.0 N The resulting solution was deep yellow and had a PH of hydrochloric acid solution. The amino acid ester hydro- 12.0. To this was added 0.38 ml. (0.003 mole) of 16.72 N chloride was recovered by evaporating the solution t o a sirup ethylenediamine solution, and the PH adjusted to 9.0 with on the steam-cone. The product was dried in a vacuum dilute nitric acid solution. The volume at this point was desiccator over phosphorus pentoxide, and purified by re- 100 ml. The solution was evaporated to about 50 ml. a t crystallization from absolute alcohol and absolute ether. room temperature and, after cooling in an ice-bath, yellow, I t is very hygroscopic and melts a t 110-112'. The yield crystalline plates of [ Pt( NOz)nen].2H20 were filtered, washed was 2.1 g. or 24%. Anal. Calcd. for CbHltOtNCl: C, with cold water, then alcohol and ether. The yield was 39.10; H, 7.86; II', 9.13. Found: C, 39.16; H, 8.02; 0.31 g., or 40% of theory. Anal. Calcd. for [Pt(NO&en]. 2Hz0: Pt, 51.00. Found: Pt, 51.25. X', 8.98. The yellow crystals were placed in an Abderhalden drying N,N-Diethylglycine hydrochloride was prepared and purified in a similar fashion. The hygroscopic, white crystals pistol a t 100' for 12 hours, after which time the product was The yield was 5.0 g., or 30%. Anal. a very pale pink in color. Anal. Calcd. for [Pt(NOz)zen]: melt a t 125-128'. Calcd. for C6H1402NCl: C, 43.00; H, 8.43; Ii, 8.36. P t , 56.30; C, 6.93; H, 2.32; Pi, 16.15. Found: P t , 56.35; C, 7.22; H, 2.36; N, 15.96. Found: C, 43.18; H, 8.39; N, 8.46. Preparation of Bis-(N,N-diethylglycine)-platinum(II), Preparation of Potassium Dinitro-( N-methyl-N-ethylgly[ Pt(amac)z] .-The freshly prepared silver oxide from 2.4 g. cine) platinate(I1) K[Pt(T\;02)2AMAC].- Kinety three (0.014 mole) of silver nitrate was vigorously shaken for ten hundredths gram (0.006 mole) of the amino acid hydrominutes with 5 ml. of water and 1.21g. (0.007 mole) of K,K- chloride was converted to the nitrate by shaking it for ten diethylglycine hydrochloride; a few glass beads were added minutes with 5 ml. of water and 100% excess silver oxide, to break up the precipitate. The precipitated silver chloride followed (after removal of the silver salts) by the addition of and excess silver oxide were filtered and washed several 60.60 ml. (0.006 mole) of 0.0984 N nitric acid. The PH of times with water. the resulting solution was 2.0. Three and ninety-seven To the filtrate was added a solution of 3.0 g. (0.007 mole) hundredths grams (0.006 mole) of Kz[Pt( N 0 2 ) 2 1 2 ] .2Hz0 was of Kn[PtCl,] in 20 ml. of water. The volume of the solution shaken one-half hour with 5 ml. of water and 100% excess was 40 ml. and its pH 4.0; this was adjusted to pH 8.5 with silver oxide. The PH of the dark yellow filtrate was 12.0. a few drops of 5% sodium hydroxide solution. The soluThis solution and that of the amino acid nitrate were mixed. tion was heated 12 hours at 70°, after which the pH was The total volume was about 100 ml. and the PH was 7.0. 4.0. The PH was adjusted to 9.0 as before and the solution A small amount of a brownish-yellow solid separated. This heated a t 70" for ten hours. Three more times, the solution became dark brown after heating on the steam-cone for was heated to 70" for two-hour intervals and the PH ad- three hours. The material was filtered and discarded, and justed t o 9.0. The final pH of the pale yellow solution was the pale yellow filtrate evaporated to dryness in a vacuum 7.0. The small deposit of platinum which had formed was over phosphorus pentoxide. The white residue was exfiltered and the fitrate concentrated t o crystallization at room tracted with several 20-ml. portions of boiling methyl alcotemperature by passing a gentle stream of air over the sur- hol, and from the filtrate white K[Pt(N02)2AMAC] was face. The white, crystalline plates were filtered on sintered

volved in this case may not be merely a preferential surface adsorption, for i t was noted that prolonged stirring of the active solutions with the quartz caused an increase in the racemization rate. Such a process could be caused by either a catalytic action by the quartz on the racemization rate, or by a difference in the rates of adsorption of the two isomers. This technique was successfully applied to 1.0% solutions of K[Pt(NO&AMAC]. Solutions which had been shaken with 1-quartz gave a positive rotation, while those that had been shaken with d-quartz gave a negative rotation. In each case, the optically active component racemized in five to six hours.

'

-

(12) I-. W Jones and

R.T.hIajor, THISJ O U R N A L , 62, 1082 (1Y30).

(13) R . N. Keller, Inorgauic Syntheses, I,247 (1946). (14) M . Vezez; A > $ ? !chirn. . p h y s . . 29, 160 (1893).

-

precipitated with absolute ether. The yield was 1.7 g., or 64%. The salt is very hygroscopic and must be stored i t 1 a desiccator. This compound could not be entirely freed from potassium nitrate, so the strychnine salt was prepared in a pure form in order to confirm the composition of the platinum complex. The impure K[Pt( NOz)zAMAC] was dissolved in 200 ml. of water and added t o 100 ml. of a solution containing 0.080 g. (0.002 mole) of strychnine nitrate. A white, flocculent precipitate formed slowly, and was filtered after cooling one hour in an ice-bath. It was recrystallized from hot water. Anal. Calcd. for (stryj[Pt(NO9)rAhlACI: Pt. 26.45: C. 42.35: 13. 4.50: N . 9.49. Found: - ~Pt,'26.35; C, 42:87;'H, 4.31; k,9.51: Resolution of K[Pt( N02)2AMAC].--The instrunlent used \vas a Schmidt and Haensch Polarimeter No. 9143. calibrated to 0.001 degree and giving reproducible results to .t0.003degree. The readings given represent an average of ten settings. (a) Resolving Agent: Quinine Bisulfate.-The resolving ;$gentwas prepared by dissolving 1.835 g. of quinine in 9.45 nil. of 0.599 M sulfuric acid and diluting to 100 ml. Twenty milliliters (0.0023 mole) of this solution was added to a solution of 1.000 g. (0.0023 mole) of K[Pt(NO&AMAC] in 100 ml. of water. The resulting solution was diluted to itbout 200 ml., and the fractions were obtained by evaporation in a vacuum desiccator over phosphorus pentoxide. Table I shows the volume of solution a t which the fractions were removed and the weight of each.

.

TABLEI Fraction

Volume when fractions removed (ml.)

Weight of fraction (g.)

125 40 15

0.151

1 2

3

,140 .442

I t was not possible to remove the adsorbed potassium bisulfate from the (quin)[Pt( NOtXAMAC] .2Hz0, but from the analytical data, the values of the elemental analyses could be calculated. Table I1 gives analytical values for the carbon, hydrogen and nitrogen compositions of the various fractions of the quinine salt, corrected for potassium bisulfate in the sample and potassium sulfate in the ash. TABLE I1 Analyses, % Hydrogen Nitrogen

Fraction

Carbon

1 2 3

39.40 39.00 39.20

4.92 5.10 4.85

9.02 9.04 9.05

(quin)[Pt(SOz)zAhLAC].aH*O 30.30

5.10

9.18

Calcd. for Portions of each of the fractions were ground for one hour in a mortar a t 0' with 2 molar equivalents of potas4um hy-

droxide solution. The precipitated quinine was filtercd and the filtrate was diluted to 10 ml. aiid examined for optical activity in a 1.0-dm. polarimeter tube. All three fractions showed levo activity and racemized upon standing. The results are summarized in Table 111.

TABLE I11 Fraction

Salt used, g.

1 2 3

0.096

Initial rotation

zero rotation ( h r . )

-0.012" ,020" .O2i0

6.25 6.00 3.00

-

.loo

.200

Time to reach

TABLEI \ ' .rype or

quartz

Run No.

lkxtro

1 2 3

Dextro Dextro Levo Levo Levo Levo

1 2

3 4

Initial rotation i P - d r n tube)

'Timc to racemize, hr.

- 0 ,024"

7.0

- ,023 - ,016

fi.5

+ +

.034 .O%X

4- .010

4- ,014

6.5 5 , (1 '1 . 5 5 ,0 5.5

In each case, both the filtrate and the residual quinine contained platinum. From this it was concluded that the (quin)(Pt(NO%),AMAC]*2H20 and quinine alkaloid u-ere somewhat similar in solubility, so that even one hour of trituration was not enough to convert all the quinine ion to the free alkaloid. However, as no precipitate formed in the filtrates, the decreasing rotation was taken as evidence of the presence of L - K [ P ~ ( N O ~ ) ~ A M A inCthe ] solution. (b) Resolving Agent: Optically Active Quartz.--The quartz powder employed was made by grinding crystals of optically active quartz in an iron mortar. The iron was removed by treatment with concentrated nitric acid followed by repeated washing with water. The particle size of the powder was,between 140 and 200 mesh. Twenty-five milliliters of a 1.0% solution of K[Pt(NO*)nAMAC] was stirred mechanically with 1.0 g. of dextroquartz for 15 minutes. The solid was filtered, 1.0 g. of fresh dextro-quartz was added and the suspension stirred for 25 minutes. The quartz was removed by filtration and the solution was found to exhibit a negative rotation which fell to zero in the course of approximately six hours. Thc ' experiment was repeated on both the same solution and a fresh one with similar results. When a 1.0% solution of K[Pt(NO-z)?AMAC] was stirred in a similar manner with 1.O g. of levo-quartz for ten minutes, the solution showed a positive rotation which fell to zero in about five hours. Repetition of the experiment with other K[Pt( S02),AMACI solutions gave similar results. Table I V summarizes the experiments. Y R R A S AJ1,i,Ix(iIs ,